1. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Lecture 42 of 42 Thursday, 03 May 2007 William H. Hsu Department of Computing and Information Sciences, KSU KSOL course page: http://snipurl.com/va60http://snipurl.com/va60 Course web site: http://www.kddresearch.org/Courses/Spring-2007/CIS560http://www.kddresearch.org/Courses/Spring-2007/CIS560 Instructor home page: http://www.cis.ksu.edu/~bhsuhttp://www.cis.ksu.edu/~bhsu Reading for Next Class: Chapter 18, Silberschatz et al., 5 th edition Data Mining, Information Retrieval and OLAP Discussion: Term Projects

2. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Chapter 18: Data Analysis and Mining Decision Support Systems Data Analysis and OLAP Data Warehousing Data Mining

3. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Decision Support Systems Decision-support systems are used to make business decisions, often based on data collected by on-line transaction- processing systems. Examples of business decisions:  What items to stock?  What insurance premium to change?  To whom to send advertisements? Examples of data used for making decisions  Retail sales transaction details  Customer profiles (income, age, gender, etc.)

4. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Cross Tabulation of sales by item-name and color The table above is an example of a cross-tabulation (cross-tab), also referred to as a pivot-table.  Values for one of the dimension attributes form the row headers  Values for another dimension attribute form the column headers  Other dimension attributes are listed on top  Values in individual cells are (aggregates of) the values of the dimension attributes that specify the cell.

5. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Data Cube A data cube is a multidimensional generalization of a cross-tab Can have n dimensions; we show 3 below Cross-tabs can be used as views on a data cube

6. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Online Analytical Processing Pivoting: changing the dimensions used in a cross-tab is called Slicing: creating a cross-tab for fixed values only  Sometimes called dicing, particularly when values for multiple dimensions are fixed. Rollup: moving from finer-granularity data to a coarser granularity Drill down: The opposite operation - that of moving from coarser-granularity data to finer-granularity data

7. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Hierarchies on Dimensions Hierarchy on dimension attributes: lets dimensions to be viewed at different levels of detail  E.g. the dimension DateTime can be used to aggregate by hour of day, date, day of week, month, quarter or year

8. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Cross Tabulation With Hierarchy Cross-tabs can be easily extended to deal with hierarchies  Can drill down or roll up on a hierarchy

9. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts OLAP Implementation The earliest OLAP systems used multidimensional arrays in memory to store data cubes, and are referred to as multidimensional OLAP (MOLAP) systems. OLAP implementations using only relational database features are called relational OLAP (ROLAP) systems Hybrid systems, which store some summaries in memory and store the base data and other summaries in a relational database, are called hybrid OLAP (HOLAP) systems.

10. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts OLAP Implementation (Cont.) Early OLAP systems precomputed all possible aggregates in order to provide online response  Space and time requirements for doing so can be very high  2 n combinations of group by  It suffices to precompute some aggregates, and compute others on demand from one of the precomputed aggregates  Can compute aggregate on (item-name, color) from an aggregate on (item- name, color, size)  For all but a few “non-decomposable” aggregates such as median  is cheaper than computing it from scratch Several optimizations available for computing multiple aggregates  Can compute aggregate on (item-name, color) from an aggregate on (item-name, color, size)  Can compute aggregates on (item-name, color, size), (item-name, color) and (item-name) using a single sorting of the base data

11. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Extended Aggregation in SQL:1999 The cube operation computes union of group by’s on every subset of the specified attributes E.g. consider the query select item-name, color, size, sum(number) from sales group by cube(item-name, color, size) This computes the union of eight different groupings of the sales relation: { (item-name, color, size), (item-name, color), (item-name, size), (color, size), (item-name), (color), (size), ( ) } where ( ) denotes an empty group by list. For each grouping, the result contains the null value for attributes not present in the grouping.

12. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Extended Aggregation (Cont.) Relational representation of cross-tab that we saw earlier, but with null in place of all, can be computed by select item-name, color, sum(number) from sales group by cube(item-name, color) The function grouping() can be applied on an attribute  Returns 1 if the value is a null value representing all, and returns 0 in all other cases. select item-name, color, size, sum(number), grouping(item-name) as item-name-flag, grouping(color) as color-flag, grouping(size) as size-flag, from sales group by cube(item-name, color, size) Can use the function decode() in the select clause to replace such nulls by a value such as all  E.g. replace item-name in first query by decode( grouping(item-name), 1, ‘all’, item-name)

13. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Extended Aggregation (Cont.) The rollup construct generates union on every prefix of specified list of attributes E.g. select item-name, color, size, sum(number) from sales group by rollup(item-name, color, size) Generates union of four groupings: { (item-name, color, size), (item-name, color), (item-name), ( ) } Rollup can be used to generate aggregates at multiple levels of a hierarchy. E.g., suppose table itemcategory(item-name, category) gives the category of each item. Then select category, item-name, sum(number) from sales, itemcategory where sales.item-name = itemcategory.item-name group by rollup(category, item-name) would give a hierarchical summary by item-name and by category.

14. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Extended Aggregation (Cont.) Multiple rollups and cubes can be used in a single group by clause  Each generates set of group by lists, cross product of sets gives overall set of group by lists E.g., select item-name, color, size, sum(number) from sales group by rollup(item-name), rollup(color, size) generates the groupings {item-name, ()} X {(color, size), (color), ()} = { (item-name, color, size), (item-name, color), (item-name), (color, size), (color), ( ) }

15. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Ranking Ranking is done in conjunction with an order by specification. Given a relation student-marks(student-id, marks) find the rank of each student. select student-id, rank( ) over (order by marks desc) as s-rank from student-marks An extra order by clause is needed to get them in sorted order select student-id, rank ( ) over (order by marks desc) as s-rank from student-marks order by s-rank Ranking may leave gaps: e.g. if 2 students have the same top mark, both have rank 1, and the next rank is 3  dense_rank does not leave gaps, so next dense rank would be 2

16. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Ranking (Cont.) Ranking can be done within partition of the data. “Find the rank of students within each section.” select student-id, section, rank ( ) over (partition by section order by marks desc) as sec-rank from student-marks, student-section where student-marks.student-id = student-section.student-id order by section, sec-rank Multiple rank clauses can occur in a single select clause Ranking is done after applying group by clause/aggregation

17. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Ranking (Cont.) Other ranking functions:  percent_rank (within partition, if partitioning is done)  cume_dist (cumulative distribution)  fraction of tuples with preceding values  row_number (non-deterministic in presence of duplicates) SQL:1999 permits the user to specify nulls first or nulls last select student-id, rank ( ) over (order by marks desc nulls last) as s-rank from student-marks

18. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Ranking (Cont.) For a given constant n, the ranking the function ntile(n) takes the tuples in each partition in the specified order, and divides them into n buckets with equal numbers of tuples. E.g.: select threetile, sum(salary) from ( select salary, ntile(3) over (order by salary) as threetile from employee) as s group by threetile

19. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Windowing Used to smooth out random variations. E.g.: moving average: “Given sales values for each date, calculate for each date the average of the sales on that day, the previous day, and the next day” Window specification in SQL:  Given relation sales(date, value) select date, sum(value) over (order by date between rows 1 preceding and 1 following) from sales Examples of other window specifications:  between rows unbounded preceding and current  rows unbounded preceding  range between 10 preceding and current row  All rows with values between current row value –10 to current value  range interval 10 day preceding  Not including current row

20. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Windowing (Cont.) Can do windowing within partitions E.g. Given a relation transaction (account-number, date-time, value), where value is positive for a deposit and negative for a withdrawal  “Find total balance of each account after each transaction on the account” select account-number, date-time, sum (value ) over (partition by account-number order by date-time rows unbounded preceding) as balance from transaction order by account-number, date-time

21. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Data Warehousing Data sources often store only current data, not historical data Corporate decision making requires a unified view of all organizational data, including historical data A data warehouse is a repository (archive) of information gathered from multiple sources, stored under a unified schema, at a single site  Greatly simplifies querying, permits study of historical trends  Shifts decision support query load away from transaction processing systems

22. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Data Warehousing

23. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Design Issues When and how to gather data  Source driven architecture: data sources transmit new information to warehouse, either continuously or periodically (e.g. at night)  Destination driven architecture: warehouse periodically requests new information from data sources  Keeping warehouse exactly synchronized with data sources (e.g. using two-phase commit) is too expensive  Usually OK to have slightly out-of-date data at warehouse  Data/updates are periodically downloaded form online transaction processing (OLTP) systems. What schema to use  Schema integration

24. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts More Warehouse Design Issues Data cleansing  E.g. correct mistakes in addresses (misspellings, zip code errors)  Merge address lists from different sources and purge duplicates How to propagate updates  Warehouse schema may be a (materialized) view of schema from data sources What data to summarize  Raw data may be too large to store on-line  Aggregate values (totals/subtotals) often suffice  Queries on raw data can often be transformed by query optimizer to use aggregate values

25. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Warehouse Schemas Dimension values are usually encoded using small integers and mapped to full values via dimension tables Resultant schema is called a star schema  More complicated schema structures  Snowflake schema: multiple levels of dimension tables  Constellation: multiple fact tables

26. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Data Warehouse Schema

27. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Data Mining Data mining is the process of semi-automatically analyzing large databases to find useful patterns Prediction based on past history  Predict if a credit card applicant poses a good credit risk, based on some attributes (income, job type, age,..) and past history  Predict if a pattern of phone calling card usage is likely to be fraudulent Some examples of prediction mechanisms:  Classification  Given a new item whose class is unknown, predict to which class it belongs  Regression formulae  Given a set of mappings for an unknown function, predict the function result for a new parameter value

28. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Data Mining (Cont.) Descriptive Patterns  Associations  Find books that are often bought by “similar” customers. If a new such customer buys one such book, suggest the others too.  Associations may be used as a first step in detecting causation  E.g. association between exposure to chemical X and cancer,  Clusters  E.g. typhoid cases were clustered in an area surrounding a contaminated well  Detection of clusters remains important in detecting epidemics

29. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Classification Rules Classification rules help assign new objects to classes.  E.g., given a new automobile insurance applicant, should he or she be classified as low risk, medium risk or high risk? Classification rules for above example could use a variety of data, such as educational level, salary, age, etc.   person P, P.degree = masters and P.income > 75,000  P.credit = excellent   person P, P.degree = bachelors and (P.income  25,000 and P.income  75,000)  P.credit = good Rules are not necessarily exact: there may be some misclassifications Classification rules can be shown compactly as a decision tree.

30. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Decision Tree

31. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Chapter 19: Information Retrieval Relevance Ranking Using Terms Relevance Using Hyperlinks Synonyms., Homonyms, and Ontologies Indexing of Documents Measuring Retrieval Effectiveness Web Search Engines Information Retrieval and Structured Data Directories

32. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Information Retrieval Systems Information retrieval (IR) systems use a simpler data model than database systems  Information organized as a collection of documents  Documents are unstructured, no schema Information retrieval locates relevant documents, on the basis of user input such as keywords or example documents  e.g., find documents containing the words “database systems” Can be used even on textual descriptions provided with non- textual data such as images Web search engines are the most familiar example of IR systems

33. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Information Retrieval Systems (Cont.) Differences from database systems  IR systems don’t deal with transactional updates (including concurrency control and recovery)  Database systems deal with structured data, with schemas that define the data organization  IR systems deal with some querying issues not generally addressed by database systems  Approximate searching by keywords  Ranking of retrieved answers by estimated degree of relevance

34. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Keyword Search In full text retrieval, all the words in each document are considered to be keywords.  We use the word term to refer to the words in a document Information-retrieval systems typically allow query expressions formed using keywords and the logical connectives and, or, and not  Ands are implicit, even if not explicitly specified Ranking of documents on the basis of estimated relevance to a query is critical  Relevance ranking is based on factors such as  Term frequency  Frequency of occurrence of query keyword in document  Inverse document frequency  How many documents the query keyword occurs in  Fewer  give more importance to keyword  Hyperlinks to documents  More links to a document  document is more important

35. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Relevance Ranking Using Terms TF-IDF (Term frequency/Inverse Document frequency) ranking:  Let n(d) = number of terms in the document d  n(d, t) = number of occurrences of term t in the document d.  Relevance of a document d to a term t  The log factor is to avoid excessive weight to frequent terms  Relevance of document to query Q n(d)n(d)n(d)n(d) n(d, t) 1 + TF (d, t) = log r (d, Q) =  TF (d, t) n(t)n(t)n(t)n(t) tQtQtQtQ

36. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Relevance Ranking Using Terms (Cont.) Most systems add to the above model  Words that occur in title, author list, section headings, etc. are given greater importance  Words whose first occurrence is late in the document are given lower importance  Very common words such as “a”, “an”, “the”, “it” etc are eliminated  Called stop words  Proximity: if keywords in query occur close together in the document, the document has higher importance than if they occur far apart Documents are returned in decreasing order of relevance score  Usually only top few documents are returned, not all

37. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Similarity Based Retrieval Similarity based retrieval - retrieve documents similar to a given document  Similarity may be defined on the basis of common words  E.g. find k terms in A with highest TF (d, t ) / n (t ) and use these terms to find relevance of other documents. Relevance feedback: Similarity can be used to refine answer set to keyword query  User selects a few relevant documents from those retrieved by keyword query, and system finds other documents similar to these Vector space model: define an n-dimensional space, where n is the number of words in the document set.  Vector for document d goes from origin to a point whose i th coordinate is TF (d,t ) / n (t )  The cosine of the angle between the vectors of two documents is used as a measure of their similarity.

38. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Relevance Using Hyperlinks Number of documents relevant to a query can be enormous if only term frequencies are taken into account Using term frequencies makes “spamming” easy  E.g. a travel agency can add many occurrences of the words “travel” to its page to make its rank very high Most of the time people are looking for pages from popular sites Idea: use popularity of Web site (e.g. how many people visit it) to rank site pages that match given keywords Problem: hard to find actual popularity of site  Solution: next slide

39. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Relevance Using Hyperlinks (Cont.) Solution: use number of hyperlinks to a site as a measure of the popularity or prestige of the site  Count only one hyperlink from each site (why? - see previous slide)  Popularity measure is for site, not for individual page  But, most hyperlinks are to root of site  Also, concept of “site” difficult to define since a URL prefix like cs.yale.edu contains many unrelated pages of varying popularity Refinements  When computing prestige based on links to a site, give more weight to links from sites that themselves have higher prestige  Definition is circular  Set up and solve system of simultaneous linear equations  Above idea is basis of the Google PageRank ranking mechanism

40. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Relevance Using Hyperlinks (Cont.) Connections to social networking theories that ranked prestige of people  E.g. the president of the U.S.A has a high prestige since many people know him  Someone known by multiple prestigious people has high prestige Hub and authority based ranking  A hub is a page that stores links to many pages (on a topic)  An authority is a page that contains actual information on a topic  Each page gets a hub prestige based on prestige of authorities that it points to  Each page gets an authority prestige based on prestige of hubs that point to it  Again, prestige definitions are cyclic, and can be got by solving linear equations  Use authority prestige when ranking answers to a query

41. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Synonyms and Homonyms Synonyms  E.g. document: “motorcycle repair”, query: “motorcycle maintenance”  need to realize that “maintenance” and “repair” are synonyms  System can extend query as “motorcycle and (repair or maintenance)” Homonyms  E.g. “object” has different meanings as noun/verb  Can disambiguate meanings (to some extent) from the context Extending queries automatically using synonyms can be problematic  Need to understand intended meaning in order to infer synonyms  Or verify synonyms with user  Synonyms may have other meanings as well

42. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Concept-Based Querying Approach  For each word, determine the concept it represents from context  Use one or more ontologies:  Hierarchical structure showing relationship between concepts  E.g.: the ISA relationship that we saw in the E-R model This approach can be used to standardize terminology in a specific field Ontologies can link multiple languages Foundation of the Semantic Web (not covered here)

43. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Indexing of Documents An inverted index maps each keyword K i to a set of documents S i that contain the keyword  Documents identified by identifiers Inverted index may record  Keyword locations within document to allow proximity based ranking  Counts of number of occurrences of keyword to compute TF and operation: Finds documents that contain all of K 1, K 2,..., K n.  Intersection S 1  S 2 .....  S n or operation: documents that contain at least one of K 1, K 2, …, K n  union, S 1  S 2 .....  S n,. Each S i is kept sorted to allow efficient intersection/union by merging  “not” can also be efficiently implemented by merging of sorted lists

44. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Measuring Retrieval Effectiveness Information-retrieval systems save space by using index structures that support only approximate retrieval. May result in:  false negative (false drop) - some relevant documents may not be retrieved.  false positive - some irrelevant documents may be retrieved.  For many applications a good index should not permit any false drops, but may permit a few false positives. Relevant performance metrics:  precision - what percentage of the retrieved documents are relevant to the query.  recall - what percentage of the documents relevant to the query were retrieved.

45. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Measuring Retrieval Effectiveness (Cont.) Recall vs. precision tradeoff:  Can increase recall by retrieving many documents (down to a low level of relevance ranking), but many irrelevant documents would be fetched, reducing precision Measures of retrieval effectiveness:  Recall as a function of number of documents fetched, or  Precision as a function of recall  Equivalently, as a function of number of documents fetched  E.g. “precision of 75% at recall of 50%, and 60% at a recall of 75%” Problem: which documents are actually relevant, and which are not

46. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Web Search Engines Web crawlers are programs that locate and gather information on the Web  Recursively follow hyperlinks present in known documents, to find other documents  Starting from a seed set of documents  Fetched documents  Handed over to an indexing system  Can be discarded after indexing, or store as a cached copy Crawling the entire Web would take a very large amount of time  Search engines typically cover only a part of the Web, not all of it  Take months to perform a single crawl

47. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Web Crawling (Cont.) Crawling is done by multiple processes on multiple machines, running in parallel  Set of links to be crawled stored in a database  New links found in crawled pages added to this set, to be crawled later Indexing process also runs on multiple machines  Creates a new copy of index instead of modifying old index  Old index is used to answer queries  After a crawl is “completed” new index becomes “old” index Multiple machines used to answer queries  Indices may be kept in memory  Queries may be routed to different machines for load balancing

48. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Information Retrieval and Structured Data Information retrieval systems originally treated documents as a collection of words Information extraction systems infer structure from documents, e.g.:  Extraction of house attributes (size, address, number of bedrooms, etc.) from a text advertisement  Extraction of topic and people named from a new article Relations or XML structures used to store extracted data  System seeks connections among data to answer queries  Question answering systems

49. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Directories Storing related documents together in a library facilitates browsing  users can see not only requested document but also related ones. Browsing is facilitated by classification system that organizes logically related documents together. Organization is hierarchical: classification hierarchy

50. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts A Classification Hierarchy For A Library System

51. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Classification DAG Documents can reside in multiple places in a hierarchy in an information retrieval system, since physical location is not important. Classification hierarchy is thus Directed Acyclic Graph (DAG)

52. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts A Classification DAG For A Library Information Retrieval System

53. Computing & Information Sciences Kansas State University Thursday, 03 May 2007CIS 560: Database System Concepts Web Directories A Web directory is just a classification directory on Web pages  E.g. Yahoo! Directory, Open Directory project  Issues:  What should the directory hierarchy be?  Given a document, which nodes of the directory are categories relevant to the document  Often done manually  Classification of documents into a hierarchy may be done based on term similarity